co-production schemes in biomass gasification · 4 process design 26 4.1 reference case 26 4.2...
TRANSCRIPT
-
Co-production schemes in biomass gasification
Citation for published version (APA):Krzelj, V. (2014). Co-production schemes in biomass gasification. Technische Universiteit Eindhoven.
Document status and date:Published: 01/01/2014
Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)
Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication
General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne
Take down policyIf you believe that this document breaches copyright please contact us at:[email protected] details and we will investigate your claim.
Download date: 07. Jul. 2021
https://pure.tue.nl/portal/en/publications/bba6d6b4-8baf-47af-b7d9-d25c46be54a1
-
Co-production schemes in biomass
gasification
Vladan Krzelj September, 2014
-
‘
Content
1 Introduction 5
1.1 BioSNG 6
2 Background 7
2.1 Project objective 8
2.2 Approach 8
2.3 Identifying the options 8
2.4 Comparison of the options 14
2.5 Preliminary economic evaluation 15
2.6 Selection 16
3 Experimental work 17
3.1 Planning of the experiments 17
3.2 Experimental setup 18
3.3 Test 1: Influence of temperature 19
3.4 Test 2: Influence of benzene addition 21
3.5 Test 3: Influence of catalyst acidity 23
3.6 Conclusions from the experimental results 25
4 Process design 26
4.1 Reference case 26
4.2 Co-production schemes 27
4.3 AspenPlus modelling 29
4.4 Results of AspenPlus simulation 30
5 Economic evaluation 31
5.1 Total Capital Investment 31
5.2 Variable costs 33
5.3 Production cost estimation 34
6 Conclusions 36
-
2
-
3
Summary
ECN is developing a technology for the production of Substitute Natural Gas
(SNG) from biomass. In the producer gas ethylene and benzene make only a
small part of the total volume (about 5 mol %) but, on the other hand, they
make a large part of the energy content. However, both of these molecules
have a higher economic value than SNG. Additionally, they are known coking
precursors for methanation catalysts. Therefore, separating them from the
producer gas may “kill two birds with one stone”.
The goal of this project is to describe and experimentally validate promising
ways to separate valuable co-products in the MILENA-OLGA to bioSNG system.
Based on differences in chemical and physical properties of benzene and
ethylene, compared to the rest of the components in the producer gas, long list
of option has been created in order to achieve project objectives. Subsequently,
it was reduced to four options: ethylbenzene production, aromatization of
ethylene, hydroformylation of ethylene and cryogenic separation.
Ethylbenzene production and aromatization of ethylene were selected as the
most promising ones from the combination of added value and expected
limited increase in capital and operating costs. Additionally, they were
considered to be most easily integrated into the current biomass to bioSNG
system. Both of those options are regarded as reactive separation techniques.
The experiments were planned and performed in order to prove the concept of
reactive separation technique. Ethylene conversion of 95% was achieved. It has
been shown, that 40% of ethylene is converted to aromatics.
Additionally, influence of temperature, addition of benzene and catalyst acidity
has been tested. It has been shown, that distribution between aromatics can be
influenced with temperature and benzene addition.
-
4
Based on the insight obtained from experimental results, coupled with
knowledge gained from literature study and following the principles of process
synthesis, two new co-production schemes in Biomass to bioSNG were created:
Ethylene benzene production process scheme (EB case)
Diluted ethylene aromatization process scheme (DEA case)
The proposed schemes were modelled in AspenPlus. The outputs from these
simulations were used for economic assessment of co-production schemes
compared to reference case.
The economic benefits of co-producing chemicals in a 100MW plant of biomass
thermal input have been quantified. It has been shown that proposed co-
production schemes are reducing production costs of SNG.
Finally, ethylbenzene production has been identified as the most promising one
offering up to 5$/GJ savings in production costs of SNG.
-
1 Introduction
Biomass is anything made from living or recently dead plant or animal material.
For thousands of years humans have harnessed natural sources of energy, such
as wind power for sailing ships and windmills, and water power from fast-
flowing rivers. We have also learned how to exploit biomass for fuel and even
today many societies still depend on wood, peat and even cow dung for fuel.
Dramatic change came with the discovery and use of fossil fuels, these being
energy dense (giving lots of energy per mass of material) and easily
transportable. It was coal that powered the Industrial Revolution and the
invention of the steam engine that revolutionized transport. Oil then followed
as an even more convenient and efficient fuel than coal, allowing humankind to
draw on vast resources of power and so increase dramatically its consumption
of energy. In the last few decades we have recognized two very stark and
significant facts. First, the supply of these fuels is finite. Secondly, by burning
these fossil fuels we are releasing vast quantities of carbon dioxide into the
atmosphere, adding significantly to the greenhouse effect and its potentially
catastrophic impact on sea levels, weather patterns and agriculture. So, a new
revolution is required to provide alternative forms of energy which are both
sustainable and kinder to the environment.
Three driving forces for the production of biofuels are:
saving a valuable finite resource - fossil fuels which will become more expensive as they become scarcer
climate change – the need to reduce the emissions of carbon dioxide
energy security - countries with a shortage of crude oil and natural gas believe that they are vulnerable and want to develop their own fuel supplies.
First generation of biofuels was based on conventional agricultural commodity
crops. This was a controversial issue because the increase in demand for
conventional food feedstocks for biofuels is likely to exacerbating price volatility
in the food market particularly the poor in developing countries. Therefore the
scientific community is changing the focus from first generation biofuels to
second generation biofuels, utilizing mainly lignocellulosic material. The main
-
6
types of first generation biofuel used commercially are biodiesel, ethanol and
biogas while the second generation biofuels are bioethanol, Fisher-Tropsch
biodiesel, Bio-DME and Bio-SNG. [1]
1.1 BioSNG
Natural gas is a convenient and environmentally friendly fuel used all over the
world. It has a wide range of applications:
Large scale electricity production in Combined Cycles
Decentralized Combined Heat and Power production (CHP)
Transportation, as a Compressed Natural Gas (CNG) or Liquefied Natural Gas (LNG)
Chemical industry, as a feedstock for many chemicals
In the Netherlands natural gas represents 46% of the Dutch (primary) energy
consumption. The main applications of natural gas are chemistry (7%), power
production (23%), and – by far the largest application – the production of heat
(70%), of which 40% is consumed by households1. However the production of
natural gas in the Netherlands has reached its maximum and it will gradually
decrease.
The substitution of natural gas by a renewable equivalent is an interesting
option to reduce the use of fossil fuels and the accompanying greenhouse gas
emissions, as well from the point of view of security of supply. In NL, the
ambition has been expressed to replace 50% of the natural gas by Green Gas
from biomass in 2050. Being a direct substitute of a natural gas, bioSNG can be
injected directly into the gas grid taking the advantage of the existing structure.
However, to be injected directly into the system, bioSNG needs to meet strict
requirement regarding Wobbe index, gross calorific value, maximum liquid
hydrocarbon content, water dew point, total sulphur content etc.
There are two main options to produce SNG from biomass. It can be done
either with anaerobic digestion (biological conversion at low temperature) or
with gasification (thermochemical conversion at high temperature).
Gasification belongs to the second generation of Bio-SNG production process
and has more potential because of the large scale and the ability to use
biomass/waste that does not interfere with food.
xxxxxxxxxxxxssssssssxxxxxxxxxxxxxx 1 Statistics Netherlands (CBS), 2006 (www.cbs.nl)
http://www.cbs.nl/
-
7
2 Background
ECN has put lot of effort into development of technology for the production of
Substitute Natural Gas (SNG) from biomass. MILENA indirect gasification and
OLGA tar removal are the first steps of the process. These technologies and
other units of the process are described in Appendix A.2. The current process
scheme is shown in the Figure 1.
MILENA OLGA HDSH2S
removal(ZnO)
Reforming
7 bar 60 bar
Biomass
Air (combustor)
Steam (riser)
SNG
H2Stars
40bar
MethanationStep 1
Steam
MethanationStep 2
Aminescrubbing
Steam
CO2
HCl removal
Figure 1: Current process scheme Biomass to bioSNG
The typical gas composition after tar removal, on dry basis, is given in the Figure
2. Ethylene and benzene make only a small part of the total volume in the
producer gas (about 5 mol %). On the other hand, as it can be seen from the
Figure 2, they make a large part of the energy content.
Both of these molecules are known coking precursors in the methanation
section. In the current system, MILENA-OLGA to SNG, ethylene is being
hydrogenated in the HDS reactor, and benzene is being reformed in the
reformer. On the other hand, both benzene and ethylene have much higher
economic value than SNG.
Figure 2: Typical gas composition after OLGA, dry basis; Share in gas heating value, after OLGA
-
8
2.1 Project objective
The goal of this project is to describe and experimentally validate promising
ways to separate valuable co-products in the MILENA-OLGA to bioSNG system.
The final objective is to quantify the possible benefits of co-production of
chemicals in a biomass to SNG system.
The main deliverables of this project are the following:
• Identifying the options from the technical and economical perspective to separate benzene and ethylene in the MILENA-OLGA to bio-SNG system.
• Selecting options and designing process scheme for overall process in the Aspen Plus software package for the 100MW plant of biomass input
• Designing and performing the experiments for validation of the important assumptions
• Quantifying the economic benefits of co-producing chemicals in a 100 MW bioSNG plant.
2.2 Approach
The approach considered for reaching the project objectives was to find the
appropriate difference in the physical or chemical properties in order to
separate ethylene and benzene from the producer gas.
2.3 Identifying the options
Due to the low partial pressure of benzene and relative high boiling point,
compared to the rest of the components present in the producer gas,
absorption would be a reasonable method for the separation of benzene. The
research about benzene, toluene and xylene (BTX) removal via scrubbing oils is
currently being performed within ECN. The first results are quite encouraging.
High level of benzene separation was achieved. [2] The technology is explained
in more details in Appendix A.2.7.
The separation of ethylene from the producer gas stream proves to be more
difficult task than separation of benzene. Two different routes were considered
for that purpose. The first is by exploiting the differences in the physical
properties such as boiling point, solubility, molecular size, etc. Several options
were considered: cryogenic distillation, membrane separation, absorption with
organic solvents and adsorption on molecular sieves. Difference in the boiling
point was considered as the only physical property that can offer a sufficient
level of separation of ethylene from the producer gas. Therefore, cryogenic
-
9
separation was considered as the only option which utilizes differences in the
physical properties.
The second route is by utilizing the chemical properties of ethylene; it’s
reactivity of the “π” bond. This approach is known as reactive separation. The
idea of this technique is to transform ethylene to products that are easier to
separate and bring additional economic value. The various reaction paths of
ethylene in chemical industry were studied in order to select possible reactions
that can be used within reactive separation method [3]. The long list of possible
reactions was created which was narrowed down eventually to three selections.
The options that were chosen, based on reactive separation approach and
studied in more details are:
Alkylation of benzene (production of ethylbenzene),
Aromatization of ethylene (mixed aromatics as a product)
Hydroformylation of ethylene (propionaldehyde as a product)
These options will be discussed in the following subchapters.
2.3.1 Alkylation of benzene (Ethylbenzene as a product)
Ethylbenzene is produced on a large scale by combining benzene and ethylene
in an acid-catalyzed chemical reaction shown in the Figure 3.
Figure 3: Alkylation of Benzene
Nearly all ethylbenzene produced in the world is used in the manufacture of
styrene; therefore, ethylbenzene demand is determined primarily by styrene
production. Consumption of ethylbenzene for uses other than the production of
styrene is estimated to be less than 1%. Current worldwide consumption of
ethylbenzene is 35 million tonnes. It is interesting to mention that
manufacturers of ethylbenzene are the major buyers of benzene, claiming more
than half of total output2.
Until 1980 aluminium chloride was used in almost all of the ethylbenzene
production facility as a catalyst (Friedel-Crafts process). Process dealt with
problems of corrosion and waste disposal. In 1980 first commercial facility using
a zeolite catalyst in vapour phase system started production (Mobil-Badger
process) which was environmentally friendly alternative to previous process.
Afterwards, liquid phase alkylation process were introduced which reduced
required benzene to ethylene ratio and decreased xylene production
(Lummus/UOP EBOne process) [4]. Liquid phase alkylation is not possible in our
system because of the excess of gases in the system (see Figure 2 ).3
xxxxxxxxxxxxssssssssxxxxxxxxxxxxxx
2 www.essentialchemicalindustry.org
3 CO,CO2, H2, CH4 are considered to be inert in the alkylation reactions
http://www.essentialchemicalindustry.org/
-
10
Typically, the reactor system in the vapour phase is comprised of the parallel
fixed bed reactors. Usual operating temperatures are between 320-450°C and
10-30 bar. Benzene is used in excess because a consecutive reaction is occurring
(ethylene is reacting with ethylbenzene producing diethyl-benzene). Choosing
the right ratio between benzene and ethylene is an optimization between
selectivity towards ethylbenzene and operating costs in the separation section.
Another reason for using benzene in excess is the temperature control since
reaction is quite exothermic. In our case ethylene is diluted so temperature
control is not considered to be an issue, since inert gases are acting as heat sink.
The catalysts used are zeolites which are noncorrosive and non-polluting. The
zeolites are essentially silica – alumina, which is environmentally inert. Because
no aqueous waste streams are produced by the process, the equipment for
waste treatment and for catalyst recovery is eliminated.
Additionally, ethylbenzene is economically more valuable product than SNG and
much easier to separate than ethylene. It can be obtained as high purity
product by distillation. The boiling point of ethylbenzene is 136⁰C compared to
-103⁰C for ethylene. Main cause for zeolite deactivation is formation of coke
inside the pores. However, the activity can be fully restored after regeneration
in the hot air stream [5]. Moreover, zeolite catalysts are considered quite
resilient to sulphur poisoning. The fact that diluted ethylene from FCC crackers
(12%) has been used in industry for ethylbenzene production makes this
approach more reasonable.
However, a disadvantage is that the stoichiometric ratio for ethylbenzene
reaction is not correct in the producer gas in MILENA-OLGA to Bio-SNG system
(4-8 to 1 in favour to ethylene). This is not considered to be a “no go” situation
since there are certain options to deal with this drawback. Different scenarios
were created in order to cope up with this challenge:
Acquiring additional benzene from an external source (petrochemical
industry- which would mean that part of the products will not be bio-
based chemicals)
Obtaining additional benzene from other SNG plants (additional three plants if biomass is used as a feedstock)
Adjusting ethylene to benzene ratio (possibility of using and developing TARA technology [6] or changing gasification conditions)
Partial hydrogenation of Ethylene (also eliminates concerns about acetylene and propylene presence in the feed but diminishes amount of product and further decreases partial pressure of the reactants)
2.3.2 Aromatization of ethylene (mixed aromatics as
products)
Oxidative coupling of methane (OCM) to ethane and ethylene has been widely
investigated in the past years. One of the drawbacks of the OCM process is the
very low concentration of ethylene in the product stream (about 5 mol%).
Separation at that low concentration is not economical. In order to make OCM
-
11
process more feasible, conversion of diluted ethylene to much less volatile
products, like aromatic hydrocarbons, has also been investigated. [7].
The challenges in OCM case, separation of diluted ethylene, are quite similar to
our case.
The idea is to make aromatics from the diluted ethylene in the producer gas
after OLGA and remove them subsequently with the BTX scrubber.
The reactions of ethylene aromatization are using zeolites as catalyst. Usually
zeolite ZSM-5 is used, due to his known low coking activity and shape
selectivity. The conversion of ethylene into aromatics involves a complex
sequence of oligomerization, isomerization, cracking and cyclization reactions
that occur on Brønsted acid sites in the zeolites. Simplified reaction path is
given in the Figure 4. Additionally it has been shown that impregnating ZSM-5
with gallium can increase selectivity to aromatics. [8]. The role of gallium is to
promote dehydrogenation of the acid-catalysed oligomerization and cyclization
products.
Figure 4: Simplified reaction path for the ethylene aromatization
Advantage of this approach is that can be easily integrated with BTX removal
process scheme. Further advantage is that product distribution can be
influenced by different temperature and space velocity. [9]
2.3.3 Hydroformylation of ethylene (Propionaldehyde as
a product)
Hydroformylation is an important commercial process for the conversion of
alkenes, carbon monoxide and hydrogen into aldehydes to be further used in
the production of various chemicals. The industrial processes are based on
homogeneous catalysis. Heterogeneous catalytic processes are currently in the
state of development [10]. In the hydroformylation reaction, olefin, carbon
monoxide, and hydrogen are reacting over a catalyst to produce an aldehyde
which has one more carbon atom than the feed olefin as it shown in the
following figure.
file:///C:/Users/KRZELJ/Desktop/heteregeneous%20Rh.pdf
-
12
Figure 5: Hydroformylation reaction The aldehyde can be treated with hydrogen to form alcohol. In commercial
operations, the hydrogenation step is usually performed immediately after the
hydroformylation step in an integrated system. The commercial catalysts are
Cobalt or Rhodium complexes. Propionaldehyde, propanol and propionic acid
are important organic chemical raw materials. The total capacity of
propionaldehyde in the world was around 260 000 t/year in 2003.
The scheme proposed for the hydroformylation case implemented in the
MILENA-OLGA to Bio-SNG would be based on LP OxoSM process with liquid
recycle (DOW chemicals, commercial process for hydroformylation of
propylene). (See Appendix A.1.1)
The main advantages for using hydroformylation as a reactive separation
technique are that product has different physical properties than ethylene
(boiling point and polarity) which makes separation easier. On the other hand,
since hydroformylation is a homogeneous catalytic reaction separation of the
catalyst is an issue. This problem can be surpassed by using ionic liquids as a
solvent [11] or by developing a process based on a heterogeneous catalysis
[10].Unfortunately, these processes in the phase of development and they are
not commercially available.
It is important to mention that Bio-SNG production is further diminished
(syngas is being used in the hydroformylation reaction) but since
propionaldehyde has a higher market value than natural gas this is considered
as a benefit. Advantage of this process is that the catalyst used (Rhodium) is
sulphur resistant. Nevertheless, it is quite expensive and since it is a
homogenous catalysis loss of catalyst is needed to be accounted for.
2.3.4 Cryogenic separation
Cryogenic separation units are operated at extremely low temperature and
often at high pressure to separate components according to their different
boiling temperatures.
In our case, producer gas would be cooled down and sent to de-methanizer.
The top fraction of de-methanizer is then feed to the system based on
Black&Veatch, Coal to LNG process. (See Appendix A.1.2). Therefore, in this
case LNG would be acquired as a product. The advantage of this approach is
that the remaining syngas (essentially H2 and CO). Therefore, it would be
reasonable to use it for a different kind of synthesis. (i.e. methanol production).
However, prior to cryogenic distillation section, gas is needed to be cleaned
from carbon dioxide and water to very low levels to avoid plugging by solidified
CO2 and water. The usual minimum specifications are 50ppm for carbon dioxide
and 1 ppm for water. This involves additional capital cost in the gas cleaning
section. Water content is reduced to low level via molecular sieve, while carbon
dioxide typically either with PSA, or specially designed amine scrubbing system
or combination of the two.
-
13
Nevertheless, the highest cost lies in the utility section for cooling down the gas
to the temperature needed for cryogenic distillation. On smaller scale, typically
mixed refrigerants are used as cooling medium. Temperature approach is
usually around 2°C in order to minimize work needed in compression section of
refrigerant cycle. In order to reduce the operating cost, process is highly
integrated which diminish flexibility for different feed compositions.
Additionally, separation of C2+ fraction is presumed not to be economical on
the small scale, due to low relative volatility between ethane, ethylene and
acetylene.
-
14
2.4 Comparison of the options
In the following table advantages and drawbacks of the short list of options
have been summarized:
Table 1: Overview of the possible options
OPTION 1
Ethyl-benzene
production
OPTION 2
Aromatization
OPTION 3
Hydroformylation
OPTION 4
Cryogenic separation
Ad
van
tage
s:
More valuable
products
Products easily
separated from the
producer gas
Easily integrated into
the Biomass-SNG
system
Zeolite catalyst: cheap
and robust
Heterogeneous
catalysis: cheap and
simple reactor
More valuable
products
Products easily
separated from the
producer gas
Zeolite catalyst:
cheap and robust
Heterogeneous
catalysis: cheap and
simple reactor
Possible high
conversion of
ethylene
Easily integrated into
the Biomass-SNG
system
More valuable
products
Product easily
separated from the
producer gas
Possibility of LNG as a
product instead of
SNG
High purity of syngas
Mature technology
Dra
wb
acks
:
Additional benzene
required
Mixed products
Homogeneous
catalysis
Difficult separation of
the catalyst
Expensive catalyst
Separation of C2+
fraction not
economically feasible
on small scale
Very low specification
of CO2 and water
Limited flexibility due
to high level of
integration
High operating costs
-
15
2.5 Preliminary economic evaluation
Rough mass balances for 100MW plant input have been done. The results are
presented in Table 3. A complete conversion of ethylene to products4 was
assumed. In Table 2, the assumed prices of possible products are given.
Table 2: The assumed prices of possible products
CH4 C2H4 Benzene EB LNG Methanol Propanal
Price [$/t] 4005 1250 1300 1400 500 400 900
Table 3: Preliminary estimation
Results in Table 3, are given for a 100MW plant assuming 8000 hours in
operation. They are used only an indication. They show price difference of the
possible products for different options in ideal case. The capital and operating
costs are not considered. Chapter 5 will describe the complete economic
picture.
xxxxxxxxxxxxssssssssxxxxxxxxxxxxxx
4 Ethylbenzene, Aromatization and Hydroformylation case
5 400 $/t equals to 7.5$/GJ LHV basis
6 Negative number represents the cost of biomass; 5 $/GJ
FEED
[kmol/hr]
SNG
case
BTX
case
EB
case
DEA
case
HF
case
CRYO
case
CH4 106 307 274 217 217 194 0
CO 246 0 0 0 0 0 0
H2 193 0 0 0 0 0 0
C2H4 35 0 0 0 0 0 35
BTX 8 0 8 -27 18 8 8
EB 0 0 0 35 0 0 0
LNG 0 0 0 0 0 0 106
Propanal 0 0 0 0 0 35 0
Methanol 0 0 0 0 0 0 146
M$/y -14 6 16 20 31 25 31 38
Difference
to ref. case
0 4 15 9 15 22
-
16
2.6 Selection
As it could be seen from the preliminary analysis, which was previously
presented in Table 3, all of the suggested options have an economic driver for
the further evaluation.
It has also been noticed that economic driver increases with the complexity of
the process. If number of equipment needed is considered as sign of the
complexity of the process, then sequence BTX, DEA, EB, HF, CRYO7 follows the
sequence of price difference of the possible products. However, the capital
costs are also expected to follow the same pattern.
However, it is important to mention that each of the alternative options will
reduce part of the capital costs of the reference case, due to the fact that two
catalytic reactors in the reference case can be omitted (HDS and reformer).
Hydroformylation case (Option 3) has a similar economic driver as EB
production case (Option 1) but since it employs a homogenous catalysis, it is
more complex process. Basically, in HF case we are solving a problem for
difficult ethylene separation but at the same time we are creating issue of
difficult separation of the catalyst. The novel processes that could overcome
this obstacle were not included since they are still in the phase of development.
Due to the high complexity and number of units needed, cryogenic separation
(Option 4) is expected to have highest capital costs and a lower chance of
success, despite the obvious benefits in differences in prices of possible
products. Nevertheless, it is important to design the process and make a
proper estimation of the investments.
Ethylbenzene production (Option 1) and aromatization of ethylene (Option 2)
were selected as the most promising ones from the combination of added value
and expected limited increase in capital and operating costs. The fact that those
options “only” involve zeolite catalysts is an important argument for selection
of these processes. It offers a possibility for a co-production to be placed
upstream since it can deal with not yet ultra clean gas. Between those options,
ethylbenzene production is a more complex one but potential economic
benefits are higher.
Therefore, both of these options were considered for the further study. Those
options were considered to have biggest chance for success from technological
and economical perspective. Furthermore, they can be easily integrated into
the current SNG process scheme while offering economic benefits.
xxxxxxxxxxxxssssssssxxxxxxxxxxxxxx
7 BTX, DEA, EB, HF, CRYO stands for BTX removal, Diluted Ethylene Aromatization, Ethylbenzene production, Hydroformylation and Cryogenic separation respectively
-
17
3 Experimental work
3.1 Planning of the experiments
As mentioned before, both of the selected options use zeolites as catalysts.
Nearly half of the patents in the field of synthesis of specialty chemicals on
zeolites concern ZSM-5, known for its low coking activity [12]. ZSM-5 based
catalysts are often used in vapour phase alkylation and offer low coking
tendency and therefore life cycles between regeneration are possible [13].
Hence, ZSM-5 was selected as a reasonable choice for performing the first
experiments.
However, in all research papers that were addressing aromatization of diluted
ethylene, benzene was not present in the gas stream. On the other hand,
alkylation of benzene with diluted ethylene was always performed with the
excess of benzene. As stated before, ethylene concentration in producer gas
exiting OLGA is usually four to eight times higher than benzene concentration. It
can be considered that conditions in our system are between these two
selected options; aromatization and alkylation. Hence, first experiment was
needed to demonstrate the behaviour of the system with these inlet feed
conditions.
Therefore, the following experiments were planned:
Test 1: Experiments with the current ratio of ethylene and benzene in the producer gas
Test 2: Experiments with the addition of benzene (stoichiometric ratio of ethylene and benzene 1:1)
These experiments were planned to be performed by using ZSM-5 with lower
Si/Al ratio. That means, that catalyst is more active, but at the same time also
more susceptible to deactivation by coking. [13]
Later on, it was planned to perform the same experiments, using ZSM-5 with
higher Si/Al ratio (Test 3).
-
18
3.2 Experimental setup
Zeolite ZSM-5 was provided by “Zeolyst International”8. Following two samples
were tested:
CBV 3014E CY (Si/Al ratio 15, 1.6 mm extrusions with 20 % binder)
CBV 8014 CY (Si/Al ratio 40, 1.6 mm extrusions with 20 % binder)
Experiments were performed in a stainless steel tubular reactor of 26mm
internal diameter, heated by electrical furnaces. 20 g of catalyst was loaded into
the reactor. The reactant and product gas composition was analysed with
microGC equipped with TCD detectors. The concentrations of lower
hydrocarbons9 and sulphur compounds were measured offline with GC set
equipped with FID detectors. The concentrations of higher aromatics10 were
measured offline with SPA11 sampling method. The conditions in reactor were
as follows: temperature= 360-450°C, pressure=1 bar, WHSV=3 h-1.
In the Figure 6 scheme of the experimental setup is given.
Figure 6: Scheme of the experimental setup
Producer gas exiting OLGA tar removal section is cooled down to 5°C which
reduces water content to around 5000ppm. Since zeolites are basically solid
acids, it is needed to purify the stream from ammonia which is considered to be
a weak base. Therefore, before entering the reactor the producer gas is cleaned
from ammonia with HNO3 solution. Neon flow of 5 ml/min was supplied to the
xxxxxxxxxxxxssssssssxxxxxxxxxxxxxx
8www.zeolyst.com
9 Propane, propylene, propadien, butane, iso-butane, C4 olefins, pentane, C5 olefins
10 The concentration of benzene and toluene were measured by micro GC
11 Solid Phase Adsorption method
-
19
system as an internal standard. Upper part of the reactor was filled with inert
alumina12 which acted as a heating zone. Temperature in the catalyst bed was
measured with set of 4 thermocouples, approximately 1cm apart in the axial
direction.
3.3 Test 1: Influence of temperature
After reaching the stable condition in the MILENA gasifier, producer gas has
been introduced to the tubular reactor. The MILENA gasifier was operating at
T= 850°C. Beechwood was used as a biomass feedstock.
The operating conditions in the tubular reactor were as follows:
Temperature=360°C, pressure=1 bar, WHSV=3h-1. Temperature was afterwards
increased to 420°C. The catalyst used was ZSM-5 with Si/Al ratio of 15. The
composition of the gas entering the reactor is given in the following table. The
values have been corrected for the air leakage that was occurring in the gas
analysis set.
Table 4: Feed composition at inlet of zeolite reactor
H2 %
CO %
CO2 %
CH4 %
N2 %
C2H4 %
C2H6 %
C2H2 %
C3H6 %
C6H6 ppm
C7H8 ppm
H2S ppm
COS ppm
21.1 35.4 21.5 11.79 3.1 4.0 0.23 0.38 0.07 5270 376 134 17
Duration of test was 4 hours. Figure 7 shows the conversion of ethylene during
that period. In the same Figure inlet and outlet concentration of benzene and
toluene are also given. The blue section on the graph represents the
temperature of 360°C while the yellow section represents the temperature of
420°C. The product distribution of ethylene products and distribution between
aromatics is given in Figure 8 .
Figure 7: Conversion of ethylene; inlet and outlet concentrations of benzene and toluene; WHSV=3h-1;
Ethylene/Benzene=7.8; ZSM-5 Si/Al=15 xxxxxxxxxxxxssssssssxxxxxxxxxxxxxx
12 99wt% αAl2O3, 0.2wt% SiO2, 3mm diameter
Be
nze
ne
an
d t
olu
en
e
con
cen
trat
ion
s [p
pm
]
0
1000
2000
3000
4000
5000
6000
7000
8000
0.00.10.20.30.40.50.60.70.80.91.0
0 50 100 150 200
con
vers
ion
of
eth
yle
ne
[%
]
Time [min]
C2H4conv
B in
B out
Tolueneout
Toluenein
-
20
Figure 8: Distribution of ethylene products; WHSV=3h-1; T=360 C; Conversion=95% E/B=7.8
The total C/H/O mass balance at 360° C gave a 97.5%, 97.4% and 97.1% match
respectively.
As it can be seen in Figure 7 initial conversion was 95%. It was decreasing over
time up to 81% where it remained stable. The initial drop or activity is assumed
to occur because of the co-adsorption of other molecules present in the gas
(mainly CO and CO2). The desorption in nitrogen flow at 470°C was performed
after the test.
It is possible that the drop in conversion is also influenced by temperature
change. However, later tests (Test 3) showed the same pattern of the
conversion drop at constant temperature.
At 360°C there was no significant change in benzene concentration. On the
other hand, the complete conversion of acetylene has been noticed. It has been
assumed that acetylene is being converted partially to ethylene and partially to
small amount of tars that have been detected (naphthalene, cresol, methyl-
naphthalene, etc.). This assumption is based on the work of Tsai and Anderson
[14] which showed that the main products of acetylene reaction over ZSM-5
zeolites, in the presence of hydrogen, are ethylene, naphthalene and methyl-
naphthalene.
From Figure 8 it can be noticed that the product of aromatics is different than in
research work of Choudary [7] and Qiu [8] which address aromatization of
diluted ethylene over ZSM-5 catalyst. In our experiment no production of
benzene has been observed at 360°C. There is also a significantly higher
concentration of ethyl-benzene and di-ethyl-benzene. Since in the inlet gas it
was already 0.5% of benzene this difference in product distribution was
expected. It can be concluded that benzene entering the reactor changes the
equilibrium between aromatic interconversion reactions. Additionally much
higher concentration of ethylbenzene and diethylbenzene are implying that
parallel reactions of benzene alkylation are occurring. As shown in the Figure 7
there was no significant difference between benzene inlet and outlet
concentration. Therefore, it can be concluded that, at 360⁰C, reaction rate of
benzene production from ethylene aromatization and reaction rate of benzene
alkylation to ethylbenzene are equal.
Aromatics 39%
Ethane 1%
Propane 19%
Propylene 5%
Propadiene
17%
C4 5%
C4= 5%
Unknown 9%
Toluene 46%
Ethylbenzene 19%
m/p-Xylene
14%
o-Xylene 3%
DEB 18%
-
21
Product distribution of ethylene converted and distribution between aromatics
were also measured at 420°C. The results are presented in the Figure 9. The total
C/H/O mass balance gave a 98.0%, 97.0% and 96.8% match respectively.
It is needed to mention that at this moment conversion was lower (81%).
Therefore, question remains if conversion has an influence on product
distribution. However, based on the work of Choudary [9] it can be assumed
that temperature has a dominant influence.
The yield of aromatics only slightly increased at 420°C. However, looking at
Figure 7 it can be observed that at higher temperature, benzene is also being
produced. This is in accordance with the findings of Choudary [9] and Qiu [8],
who showed that at higher temperature the product distribution is shifted
towards lower aromatics.
Figure 9: Distribution of ethylene products;
WHSV=3h-1; T=420C; ZSM 5 Si/Al=15 ;Conversion =81%; E/B=7.8;
3.4 Test 2: Influence of benzene addition
The second test was planned to be performed with stoichiometric ratio of
ethylene and benzene in the producer gas entering the reactor. Increasing the
benzene concentration was planned to be done by feeding toluene into the
MILENA gasifier. The main reason for taking this approach was due to safety
issues. After reaching stable condition in MILENA gasifier toluene was feed to
the gasifier with the mass flow of 200g/h and subsequently 600 g/h. However,
the maximum increase of benzene concentration that was achievable with this
approach was 0.6 mol% (total 1.1 mol%). This is far from the concentration of
benzene needed for the stoichiometric ratio between ethylene and benzene (4
mol%).
Benzene 12%
Toluene 45% Ethylbenze
ne 11%
m/p-Xylene 15%
o-Xylene 3%
DEB 14%
Aromatics 41%
Ethane 1%
Propane 17%
Propylene 17%
Propadiene 10%
C4 4% C4=
10%
-
22
Figure 10: Concetration of benzene and toluene in OLGA gas
The yellow section in the figure above represents the beginning of toluene
feeding into the gasifier of 200g/h while the green section represents the
increase of toluene feeding to 600 g/h.
Total mass C/H/O balance for MILENA gasifier did not give a satisfactory match.
However, increase of CO2 concentration of the flue gas in the combustion
sector was clearly perceived. It is shown in the Figure 11. It indicates that
additional carbon (from toluene) is transported to the combustion section and
therefore is lost from the producer gas.
Figure 11: Concetration of CO2 and O2 in the flue gas
.
However, although desirable concentration in the inlet was not achieved, it was
noticed that even small increase of benzene from 0.5% to 1.1% would increase
aromatic distribution towards ethyl-benzene and di-ethyl-benzene significantly.
0
500
1000
1500
2000
2500
3000
0
2000
4000
6000
8000
10000
12000
8:00 8:30 9:00 9:30 10:00 10:30 11:00 11:30 12:00 12:30
Be
nze
ne
co
nce
trat
ion
, pp
m
Time
BenzeneToluene
0
2
4
6
8
10
12
8:00 9:00 10:00 11:00 12:00 13:00
Co
nce
trat
ion
vo
l%
Time
O2 %
CO2 %
Tolu
en
e c
on
cen
trat
ion
, pp
m
-
23
The conversion of ethylene was 87 %, of which product distribution towards
aromatics was 37%. The product distribution between aromatics is shown in the
Figure 12.
Figure 12: Distribution of aromatics products in case of benzene addition; WHSV=3h-1;T=360°C; ZSM-5
Si/Al=15 ; Benzene inlet = 1.1 mol%
Nevertheless, it is import to mention that toluene inlet concentration was
increased as well, which would diminish toluene production in aromatization
reactions due to equilibrium existing between the aromatic products.
Therefore, it remains unknown what would be product distribution towards
toluene in case when only benzene concentration is increased.
3.5 Test 3: Influence of catalyst acidity
Additional test was performed to test the stability of conversion during the
longer period of time. Catalyst used in this case was ZSM with Si/Al ratio of 40.
The composition of inlet feed to the reactor is given in Table 5.
Table 5: Feed composition at inlet of zeolite reactor
H2 %
CO %
CO2%
CH4 %
N2 %
C2H4%
C2H6%
C2H2 %
C3H6 %
C6H6 ppm
C7H8 ppm
H2S ppm
COS ppm
28 27.7 27.8 10.2 2.6 2.8 0.13 0.19 0.03 5526 239 118 10
The conversion of ethylene and benzene during 6h test are given in Figure 13.
EB 47%
Xylenes 11%
DEB 42%
-
24
.
Figure 13: Conversion of ethylene and benzene; WHSV=3h-1; T=360C; ZSM-5 Si/Al=40
It has been noticed that initial conversion was lower than in previous cases
(45% compared to 95%) as well as stable conversion (20% compared to 81%).
That was expected, since zeolites with higher Si/Al ratio have lower acid site
density, which implies lower activity. However, it must be stressed, that inlet
conditions were not the same as in the previous cases. It can be noticed from
Table 5, that ethylene inlet concentration was 31% lower than in the previous
tests. Since reaction rate is dependent of the partial pressure of ethylene, lower
inlet concentrations imply lower conversion. The sudden drop of ethylene was
due to disturbance in the flow in the gas analysis system.
Additionally, during this test benzene conversion of approximately 50% was
observed. The possible reasons are:
• Higher Si/Al ratio favours the alkylation reactions
• Lower E/B ratio favours the ethylbenzene production
• Production of ethylbenzene is favoured at lower conversion of ethylene
However, additional experiments are needed to be performed in order to
distinguish what is the dominant factor for this observed behaviour.
Unfortunately, until now SPA samples are not being analysed, so distribution of
aromatic products is not given in this version of the report.
4.0
4.5
5.0
5.5
6.0
0
0.1
0.2
0.3
0.4
0.5
0.6
0 100 200 300 400
Co
nve
rsio
n
Time [min]
Benzeneconversion
Ethyleneconversion
E/B ratio
E/B
rat
io
-
25
3.6 Conclusions from the experimental results
The following conclusions have been made from the experimental work:
• The idea of reactive separation was proven
• High conversion of ethylene is achievable
• Selectivity towards aromatics was around 40%
• Acetylene is fully converted
• High temperature favours the production of lower aromatics
• Higher concentration of benzene at inlet are shifting the product distribution towards ethylbenzene
• Feeding toluene into the gasifier can increase benzene concentration in producer gas only up to the certain point
However, main insight of the experimental work performed was that reactive
separation of diluted ethylene was proven as a concept. Ethylene as a small
molecule with low boiling point is converted to large aromatic molecules with
high boiling point.
The important aspect of the performed experiments is also that it gave insight
for designing the process schemes, and background to support currently
theoretical co-production schemes. In order to make co-production schemes
more feasible it would be desirable to have high selectivity towards aromatics.
Also high selectivity toward one aromatic product would be desirable, because
of the higher price that we can get for pure products.
The certain conclusions have been made that showed the way to achieve it.
Even though we did not manage to achieve the conditions when benzene is in
excess compared to ethylene, there is a clear indication that additional benzene
would increase selectivity towards ethylbenzene significantly.
Therefore, two possible routes are possible:
Developing the catalyst in order to achieve high selectivity towards aromatics and optimizing reactor conditions in order to increase selectivity towards one product.
Supplying the additional benzene to the system in order to increase selectivity towards one product (ethylbenzene)
-
26
MILENA OLGA HDSH2S removal
(ZnO)Reforming
7 bar 60 bar
Biomass
Air (combustor)
Steam (riser)
SNG
H2Stars
40bar
MethanationStep 1
Steam
MethanationStep 2
Aminescrubbing
Steam
CO2
Cl removal
4 Process design
In the following subchapters process scheme of the current Biomass-bioSNG
system is given with a short description. Gradually, the alternative co-
production schemes are being introduced and described briefly. The newly
developed process schemes ethylbenzene production (EB case) and diluted
ethylene aromatization (DEA case) are presented in Chapter 4.2.2. and Chapter
4.2.3. Their development was based on literature study, insights gained from
the lab experiments and principles of process synthesis. The unit used in all of
the process schemes are described in more details in Appendix A.2.
4.1 Reference case
Figure 14: Reference case process scheme
The reference process considers that SNG is the only product made from
biomass gasification. The block flow diagram of the process is shown in Figure
14. Here we can see that after gasification and tar removal, the gas is cooled to
25 oC. and further compressed to a pressure of 7 bar. After compression some
water will condense out from the gas stream, while the gas at high
temperature, obtained in the last compression stage, is sent to an adsorption
unit where HCl is retained on ZnO+Na2O/Al2O3 adsorbent at 180 oC such that in
the outlet the composition of HCl is 1 ppm (mol) or lower. After the HCl
removal, the gas is sent to hydrodesulphurisation (HDS) unit where organic-S is
converted to H2S. At the same time, it is considered that HDS catalyst has shift
and hydrogenation activity of unsaturated hydrocarbons (benzene is not
hydrogenated). In the following step, H2S is retained by chemical adsorption on
-
27
ZnO and the product gas free of H2S is sent to a reforming unit, where heavy
hydrocarbons (e.g. benzene, toluene) are reformed into smaller molecules (CH4,
CO, CO2, H2). In addition, water gas shift (WGS) also occurs in the reforming. The
product of the pre-reformer is sent to the first methanation step, since it is free
of H2S and other components that may deactivate the methanation catalyst
(HCl, benzene). At the same time, the feed to 1st methanation step is at high
temperature and contains a significant amount of water to prevent soot
formation, and this is an advantage compared to the case when 1st methanation
would be carried out after CO2 separation. After the first methanation step, CO2
is removed by amine scrubbing. The amount of CO2 removed is such that
stoichiometric ratio between the components (CO, CO2, H2) participating in the
methanation is at stoichiometric value. After the last methanation step, the
SNG product is compressed to 60 bar.
4.2 Co-production schemes
4.2.1 BTX removal case
Figure 15: BTX removal process scheme
Compared to the reference case, in this process layout the benzene and other
heavy organic components are removed from the stream leaving the tar
removal unit (see Figure 14). Benzene separation is carried out by washing it out
with a solvent. The product stream, free of benzene is cooled to 25⁰C to
condense out the water, and further sent through an active carbon bed to
retain all the heavy components that could not be washed out with benzene.
After compression to 7 bar, the HCl is removed by adsorption similar as in the
reference case. The product stream free of HCl is sent to a WGS reactor, and for
this purposes some steam is added. Besides WGS, in the same reactor
hydrogenation of unsaturated hydrocarbons (ethylene) occurs. One difference
with the reference case is that WGS and hydrogenation takes place in the
presence of H2S, which is possible on Co/Mo-S catalyst. Due to this, the first
methanation step is no longer in front of CO2 separation. However, the
advantage of this process is that H2S can be removed together with CO2 in the
amine-scrubbing unit minimizing the use of ZnO for H2S separation. After CO2
removal and compression to 40 bar still a guard bed is needed to protect the
methanation catalyst. After the methanation, the SNG product is compressed to
the imposed pressure of 60 bar.
MILENA OLGABTX
removalSour-WGS
Aminescrubbing
Methanation
7bar 40 bar 60 bar
Biomass
Air (combustor)
Steam(riser)
SNG
Steam
BTX CO2+H2S
tars
Steam
Cl removal
-
28
4.2.2 EB case
Figure 16: EB case process scheme
Compared to the BTX removal case (see Figure 15), the difference is that
producer gas leaving tar removal section is sent first to the ammonia removal
unit. NH3 needs to be removed prior to zeolite catalyst. Because of the HCl and
CO2 present in the producer gas simple water scrubbing can be applied for that
purpose. HCl is being removed as well so there is no need later on for a HCL
adsorbent.
Gas is then compressed to 7 bar. Before the reactor, producer gas is mixed with
benzene, in order to reach benzene to ethylene ratio of five. In the alkylation
reactor, ethylene reacts with benzene over zeolite catalyst and produces
ethylbenzene. Afterwards, excess of benzene and alkylation products are
separated from the gas stream in the separation section (see appendix A.2.5).
Separation section in the EB case consists of flash drum, distillation column and
BTX scrubber. Ethylbenzene is taken as a product while benzene is recycled
back to the alkylation reactor. Benzene from outside source is also added to the
recycle stream. The gas stream is heading next to the Sour-WGS reactor. The
difference in this case is that only small fraction of unconverted ethylene is
being hydrogenated. That means more hydrogen in producer gas, which further
implies less steam needed in the WGS reaction and less CO2 removed in amine
scrubber. In this case H2S is also removed in the amine scrubbing unit. The
producer gas is then sent to the methanation section. After the methanation,
the SNG product is compressed to the imposed pressure of 60 bar.
4.2.3 DEA case
Figure 17:DEA case process scheme
MILENA OLGAEB
reactorSour-WGS
Aminescrubbing
Methanation
7bar 40 bar 60 bar
Biomass
Air(combustor)
Steam(riser)
SNG
Steam
CO2+H2Stars
Separation sector
EB
Benzene recycleBenzeneSteam
Waterscrubbing
NH3+HCl
MILENA OLGA DEA reactor Sour-WGSAmine
scrubingMethanation
7bar 40 bar 60 bar
Biomass
Air (combustor)
Steam (riser)
SNG
Steam
CO2+H2S
tars
Aromaticsremoval
Aromatics
Steam
Waterscrubbing
NH3+HCl
-
29
This case can be considered as simplified version of EB case, previously
described. Difference is that no benzene is supplied to the system and therefore
there is no need for complicated separation system for benzene recycle.
After the DEA reactor, aromatics are removed via scrubbing oil. The gas stream
is heading next to Sour-WGS reactor. Steam is added and WGS reaction is
occurring. In the same reactor unconverted ethylene and lower unsaturated
hydrocarbons produced are being hydrogenated. Subsequently, producer gas is
feed to the amine scrubber where H2S and excess of CO2 are removed. At this
point producer gas is cleaned from olefins, aromatics, higher compounds and
H2S and it has a right M ratio for the methanation reaction. The gas is
compressed to 40 bar and sent to the methanation section. After the
methanation, the SNG product is compressed to the imposed pressure of 60
bar.
The short overview of the differences between all the options described is given
in Appendix 0.
4.3 AspenPlus modelling
AspenPlus 8.2 has been used for the flowsheet calculations, with SRK chosen as
the thermodynamic model and the steam tables obtained from STEAMNBS. The
volume percentages of feed (OLGA gas) were the same as in the reference case.
The PFDs and heat & mass balances can be found in Appendix 0.
The following three cases were assessed:
EB: Ethylbenzene production case (see Figure 16) o EB-1: Conversion of ethylene and selectivity to ethylbenzene
obtained from the literature [15]
DEA: Aromatization of diluted ethylene (see Figure 17) o DEA-1: Conversion of ethylene and distribution of the products
obtained from the lab experiments o DEA-2: Conversion of ethylene selectivity to aromatics products
obtained from the literature [8] Distribution between aromatics obtained from the lab experiments
General descriptions of the units used in all process schemes and assumption
that have been made in AspenPlus model are given in Appendix A.2 .
-
30
4.4 Results of AspenPlus simulation
The results of the AspenPlus simulations are given in Table 6. In the same table,
in the first two columns, the simulation results from A. Motelica [16] for the
Reference and BTX removal case are also given.
These results will be used as an input for the cost estimate and comparison in
the following chapter.
Table 6: The results of the AspenPlus simulations
SNG -
Base
case
BTX
case
EB -1
case
DEA-1
case
DEA-2
case
SNG [kg/hr] 5421 4838 3998 4508 4258
SNG
composition
[mol%]
CH4
H2
CO2
91.8
3.7
0.9
91.8
3.7
0.9
90.7
3.5
1.4
92.1
2.7
1.4
91.1
3.4
1.4
LHV [MJ/kg] 46.1 46 44.8 45.2 44.9
Co-products
[kg/hr]
679 3660 1026 1438
Co-product composition
[mol%]
B
T
91.0
8.9
EB
DEB
B
T
86.7
10.0
1.2
2.1
B
T
X
EB
DEB
68.8
19.3
4.8
2.9
2.9
B
T
X
EB
DEB
54.2
26.8
8.2
4.9
4.9
Benzene added [kg/hr] -2032
Steam available [kg/hr] 6420 5102 2815 7708 8507
Electricity [MW]
(compressors and pumps)
3.63 3.55 3.52 3.50 3.49
CO2 removed [kg/hr] 13081 12267 11309 11689 11268
B: Benzene
T: Toluene
X: Xylene
EB: Ethylbenzene
DEB: Diethylbenzene
-
31
5 Economic evaluation
The cost comparison of reference case with alternative process schemes has
been done. The effects of co-production schemes on Total Capital Investment
and Variable cost have been evaluated in the following subchapters.
5.1 Total Capital Investment
The Total Capital Investment constitutes multiple factors, as shown in Table 7.
Typical breakdown of costs for a similar type of plant is also given. [17]
Table 7: Average breakdown of Total Capital Investment, as used for this study
TOTAL CAPITAL INVESTMENT
Fixed Capital Investment
Working
Capital
Start-up
Costs Direct Costs Indirect Costs
ISBL costs (on-site)
OSBL Costs (off-site)
Purchase & Installation of
Process equipment
Piping & appurtenances
Instrumentation & Controls
Electric equipment & materials
Civil & structural
Process Buildings
Yard Improvements
Auxiliary buildings
Service facilities
Storage & distribution
Land
Up-front R&D
Up-front license
Equipment design and engineering
Construction
Contractor’s fee
Contingencies
Inventories
Salaries/wages due
Receivables less payables
Cash
Modifications
Start-up labour
Loss in production
45% of TCI 20% of TCI 18% of TCI 10% of TCI 7% of TCI
100% 44% of ISBL 40% of ISBL 22% of ISBL 16% of ISBL
-
32
There are two approaches for estimating Total Capital Investment of a bioSNG
plant:
From reference data of technologically similar facilities, applying different factors to cope with different scale, start-up year, etc.
Bottom up method, where basic engineering is carried out for the plant and cost estimation are based on process, flow diagram, energy and mass balances, and quotes for major equipment.
In the report from G. Aranda [18] the first method was pursued in order to get a
first estimation of the costs for bioSNG plant. Total capital investment was
estimated to be 1100$2013/kWinput. This value has been estimated for a 1GW
scale of input. Scaled down to 100 MW (input biomass for AspenPlus
simulations) it gives Total Capital Investment cost of 220 M$.
It can be assumed that co-production does not have an effect on working
capital cost and start-up costs. Regarding the OSBL costs (off-site), small
increase in costs can occur because of the storage capacity needed for the co-
products. However, due to the small scale of the plant considered in this study
it is expected to be insignificant.
Indirect costs include the construction of the plant, contractor's fees, etc. These
also include R&D, design and engineering of the equipment, and licenses.
Comparing the co-production schemes, we can assume that only EB case could
have higher indirect cost than reference case. This is due to the fact that EB
case is more complex due to the separation sector. For other two alternatives, it
is anticipated to have same indirect cost as the reference case.
If we consider breakdown of the TCI for a 100 MW, according to Table 7 , we are getting ISBL costs of 99M$. Deducting from that amount ,the typical costs
for piping, civil and structural, instrumentation and control, gives cost of 88M$
for purchase and installation of all process equipment. MILENA and OLGA are
considered to make 40% of those costs13. Therefore, since modifications of our
system are located downstream of OLGA, it can be concluded that alternative
schemes are not changing the TCI significantly.
This is even more expressed when impact of added equipment on rest of the
process is taken into consideration. For example, adding equipment like BTX
scrubber will remove other unit like reformer and HDS. Or else, removing
ethylene and benzene from the stream, would diminish SNG production, but at
the same time methanation section would be smaller and, therefore, cheaper.
Additionally, both of the zeolite reactors in the proposed schemes are fixed bed
reactors; therefore it is not expected to contribute significantly to the total
equipment costs.
Finally, it can be concluded that alternative co-production schemes proposed
are not expecting to have a significant influence on TCI; except in EB case
where minor additional TCI cost are anticipated due to the separation sector for
benzene recycle.
xxxxxxxxxxxxssssssssxxxxxxxxxxxxxx
13 Internal data
-
33
5.2 Variable costs
The comparison of variable costs was based on the results from the Aspen
simulation and certain assumptions. The results from the evaluation are
presented in Table 8. The prices of the catalysts and assumptions that had been
made are given in Table 9.
Table 8: Variable costs for all cases in k$/y
K$/y SNG BTX EB DEA-1 DEA-2
Cost of ZSM5 0 0 403 341 0
Cost of GA/ZSM5 0 0 0 0 504
Cost of reformer catalyst 1834 0 0 0 0
Cost of electricity compressors 2262 2222 2216 2191 2212
Cost of electricity amine scrubber 202 189 175 180 174
Cost of electricity BTX removal 150 150 150 150
Cost of Cl adsorbent 19 19 0 0 0
Cost of H2S adsorbent 79 0 0 0 0
Cost of make-up BTX scrubber 0 4 4 4 4
Cost of make-up amines 20 20 20 20 20
Sum of operating cost 4416 2604 2968 2886 3064
From the Table 8 it can be realized that all of the alternative schemes have a
noticeably lower operating costs. The main reason for the higher operating
costs in the reference case is due to the reformer catalyst cost. Difference in
operating cost between two DEA cases is because DEA-2 case uses zeolite
catalyst with 5% impregnated gallium in order to achieve higher selectivity. The
reason for assuming different weight hourly space velocities for EB and DEA
cases was due to the fact that reaction rate of alkylation reaction increases with
higher partial pressure of benzene.
Table 9: Prices and assumption used for variable cost estimate
Price Assumption used:
ZSM 5 60 $/kg
EB 1: WHSV=3h-1, two reactors; catalyst lifetime of 3 years
DEA 1 : WHSV=2h-1, two reactors; catalyst lifetime 3 years
ZSM/Ga 5wt% 89 $/kg DEA 2 : WHSV=2h-1, two reactors; Ga price 280$/kg; lifetime of 3 years
Reformer catalyst 195 $/kg NSHV=500; Bulk density=750 kg/m3; catalyst lifetime of 3 years
Cl and S adsorbent 170 $/ton 13% adsorbing capacity
Electricity 0.075 $/kWh
-
34
5.3 Production cost estimation
In the following table the final result of cost evaluation are given.
k$/y SNG BTX EB DEA-1 DEA-2
TCI annually 22000 22000 22500 22000 22000
Operating costs 4416 2604 2968 2886 3064
Fixed 3300 3300 3375 3300 3300
Biomass 14400 14400 14400 14400 14400
Total cost 44116 42304 43243 42586 42764
BTX -5974 17880
AROM -9029 -12651
EB -36605
SNG produced [PJ/y] 2.00 1.78 1.43 1.63 1.53
Cost SNG [$/GJ] 22.1 20.4 17.1 20.6 19.7
Following assumption were made in the cost evaluation presented above:
Interest and depreciation 10 % of the total capital investment14
Operating hours: 8000h
Biomass cost: 5 $/GJ
BTX price: 1100 $/tonne
Ethylbenzene price: 1250$/tonne
Mixed aromatics price: 1100$/tonne
5 M$ higher TCI for EB case15
It can be seen that all of the co-production schemes offer a reduction in
production costs of bioSNG.
It has been shown that EB case is a favourable one, offering a reduction in
production cost of 5$/GJ, even when additional TCI costs are accounted. On the
other hand, results of the economic analysis depend significantly on the
assumed prices of the co-products.
Ethylbenzene prices are not easily found in open literature. However, as can be
seen in Appendix A.5.1, price of styrene is mainly influenced by benzene price.
The average price difference between styrene and benzene is around
250$/tonne up to 300$/tonne. Limited data found in literature16 showed that
that ethylbenzene to benzene price difference is around 100-200$ per tonne.
Therefore, it can be considered that assumed prices used in this economic
xxxxxxxxxxxxssssssssxxxxxxxxxxxxxx
14 Due to simplicity reasons; It is equal to an interest rate of 5.5% and depreciation period of 15 years
15 It is a combination of cost analysis of the distillation column in Aspen-Icarus and Indirect costs of 18% TCI
16 www.icis.com
-
35
assessment were reasonable. The influence of price difference between
ethylbenzene and benzene on SNG production costs is also presented in
Appendix A.5.1.
DEA-1, which used first results from lab experiments as an input, show to be
less feasible than BTX removal. Still, evaluation of DEA-1 case was performed
with first lab results, which were not optimized.
DEA-2 case proved to offer more cost benefit than BTX case. However, price
considered for mixed aromatic products was the same as price assumed for co-
products in BTX case (91% benzene, 9% toluene).
In Figure 18 influence of co-product price on SNG production costs is presented.
In same figure, for EB case it is assumed that price difference between
ethylbenzene and benzene is 150 $/tonne.
Figure 18: Influence of co-product prices on SNG production costs; Biomass=5$/GJ
As it can be seen from Figure 18, co-production schemes offer the savings in
production costs of SNG, even when low prices are assumed for co-products.
Nevertheless, price of the biomass has also a lot of influence on feasibility of co-
production schemes, particularly when acquired prices of co-products are low.
(see Appendix A.5.2)
15.0
16.0
17.0
18.0
19.0
20.0
21.0
22.0
23.0
800 900 1000 1100 1200 1300
SNG
pro
du
ctio
n c
ost
[$/G
J]
Price of co-produts [$/tonne]
REFERENCE
BTX
DEA-1
DEA-2
EB
-
36
6 Conclusions
Ethylbenzene production and aromatization of ethylene were selected as the
most promising options for separation of valuable co-products in the MILENA-
OLGA to bioSNG system. They were selected as the best combination of added
value and expected limited increase in capital and operating costs. Additionally,
they were considered to be most easily integrated into the current biomass to
bioSNG system. The fact that those options “only” involve zeolites, offers a
possibility for a co-production to be placed upstream since it can deal with not
yet ultra clean gas.
The experiments were planned and performed in order to prove the concept of
reactive separation technique. Ethylene conversion of 95% was achieved. It has
been shown, that 40% of ethylene is converted to aromatics.
Additionally, influence of temperature, addition of benzene and catalyst acidity
has been tested. It has been shown, that distribution between aromatics can be
influenced with temperature and benzene addition.
Based on the insight obtained from experimental results, coupled with
knowledge gained from literature study and following the principles of process
synthesis, two new co-production schemes in Biomass to bioSNG were created:
Ethylene benzene production process scheme (EB case)
Diluted ethylene aromatization process scheme (DEA case)
The economic benefits of co-producing chemicals in a 100MW plant of biomass
thermal input have been quantified. It was shown that their influence on TCI is
limited while at the same time co-productions schemes are reducing operating
costs mainly because catalytic reformer is omitted.
Ethylbenzene production case was selected as favourable option with the
possibility to reduce production cost of bioSNG up to 5$/GJ.
-
37
It was shown that all of the co-production schemes offer a reduction in SNG
production costs. However, reduction in production costs varies significantly
with prices of co-products and price of biomass.
Nevertheless, they offer a substantiate reasoning to evaluate the proposed co-
production schemes in more details. Furthermore, promising results from the
first experiments offer a solid backbone that those options are plausible.
Therefore, they offer a potential for future research. Further development is
needed, optimization of catalyst and operating condition. Longer experiments
need to be performed in order to assess catalyst lifetime between regeneration.
Finally, the results of this research were only one of the first steps in achieving
the co-production of bio-chemicals in Biomass to SNG system. However, the
one can imagine that after a persevering development of co-production of
biochemical, it can make bioSNG more competitive to other energy sources.
-
38
Bibliography
[1] S. Naik, "Production of first and second generation biofuels: A comprehesive
review," Renewable and Sustainable Energy Reviews, pp. 578-597, 2010.
[2] A. Bos, M. Cieplik, A. van der Drift, B. Vreugdenhil and J. Kuipers, "BioBTX
Ontwikkeling van een proces voor verwijdering van benzeen, tolueen en xyleen,"
ECN, 2013.
[3] S. Matar and L. F. Hatch, Chemistry of Petrochemical Processes, 2nd edition, 2000.
[4] C. Perego and P. Ingalina, "Recent advances in the industrial alkylation of
aromatics," Catalysis today, pp. 3-22, 2002.
[5] Y. Song and S. Liu, "Coke burning behaviour of a catalyst of ZSM-5/ZSM-11 in the
alkylation of benzene with FCC off-gas to ethylbenzene," Fuel Processing
Technology, pp. 297-302, 2006.
[6] M. Carbo, B. Vreugdenhil and L. Bleijendaal, "Test resultaten lab-schaal TARA-
reactor," ECN, 2012.
[7] V. Choudary and P. Devadas, "Aromatization of dilute ethylene over Ga-modified
ZSM-5 type zeolite catalysts," Microporous and Mesoporous Materials, pp. 253-
267, 2001.
[8] P. Qiu, J. Lunsford and M. P. Rosynek, "Characterization of Ga/ZSM-5 for the
catalytic aromatization of dilute ethylene streams," Catalysis Letters, pp. 37-42,
1998.
[9] V. Choudary, S. Banerjee and D. Panjala, "Influence of Temperature on the Product
Selectivity and Distribution of Aromatics and C8 Isomers in the Conversion of Dilute
Ethene over H-Galloaluminosilicate (ZSM-5 type) Zeolite," Journal of Catalysis, pp.
398-403, 2002.
[10] T. Zeelie, "Rhodium and Cobalt catalysts in the heterogeneous hydroformylation of
ethene, propene and 1-hexene," Industrial Chemistry Publication Series, 2007.
[11] Y. Diao, L. Jing and L. Wang, "Ethylene hydroformylation in imidazolium-based ionic
liquids catalyzed by Rhodium-phosphine complexes," Catalysis today, pp. 54-62,
2013.
[12] G. Perot and M. Guisnet, "Advantages and disadvantages of zeolites as catalysts in
organic chemistry," Journal of Molecular Catalysis, pp. 173-196, 1990.
-
39
[13] N. Chen and W. Garwood, "Catalysis Review," Science and Engineering, pp. 185-
264, 1986.
[14] P. Tsai and J. Anderson, "Reaction of Acetylene over ZSM-5 Type Catalysts," Journal
of Catalysis, pp. 207-214, 1983.
[15] P. Graf and L. Lefferts, "Reactive separation of ethylene from the effluent gas of
methane oxidative coupling via alkylation of benzene to ethylbenzene on ZSM-5,"
Chemical Engineering Science, pp. 2773-2780, 2009.
[16] A. Motelica, "BTX removal in biomass processing," ECN, 2013.
[17] M. Carbo and R. Smit, "SNG production for GdF SUEZ," ECN.
[18] G. Aranda, A. van der Drift and R. Smit, "The Economy of Large Scale Biomass to
Substitute Natural Gas (bioSNG) plants," ECN, 2014.
[19] N. Hansen and R. Krishna, "Reactor simulation of benzene ethylation and ethane
dehydrogenation catalyzed by ZSM-5: A multiscale approach".
[20] M. Abu-Zahra, J. Niederer and P. Feron, "CO2 capture from power plants; A
parametric study of the economic perfomance basd on mono-ethanolamine,"
International journal of greenhouse gas control, pp. 135-142, 2007.
[21] R. J., "Science and Development Aspects of the Haldor Topsoe TREMP Methanation
Technology," 2007.
-
40
Appendix A. Options
A.1. Co-production options
A.1.1 Hydroformylation case
Figure 19:Process scheme of Oxo process n-butyraldehyde production process
-
41
A.1.2 Cryogenic case
Figure 20: Process scheme of Black&Veatch Coal to LNG process
-
42
A.2. Description of units in the process schemes
A.2.1 MILENA gasifier (all cases)
MILENA is an allothermal gasifier developed by the Energy Centre of the
Netherlands (ECN). Its development started in the late 1990’s based on the
experiences of the SilvaGas and the FICFB (Güssing) 17technologies. Result was
lab-scale MILENA in 2004 of approximately 25KWth input capacity and a pilot-
scale place in 2008 of 800KWth capacity. Both of the gasifiers are connected
with OLGA technology for tar removal.
It was designed as a first step in the production of synthetic natural gas (SNG).
The gasifier contains separate sections for gasification and combustion. The
gasification section consists of three parts (gasifier riser, settling chamber and
downcomer) while combustion chamber consists only of one part. A scheme of
the MILENA technology is shown in the following figure and it will be explained
subsequently.
Biomass is fed into the gasifier riser. A small amount of superheated steam is
added from below. Steam can be replaced by air if dilution of the producer gas
is not a problem. Hot bed material (typically at 925⁰C) enters the riser through a
hole, opposite of the biomass feeding point. The bed material heats the
biomass to 850⁰C in the gasification section. The heated biomass particles
degasify and are partially converted into gas. The bed material, together with
the char is carried through the riser and separated from the producer gas in the
settling chamber falling down into the downcomer and carried to the
combustor. The typical residence time of the gas is several seconds.
xxxxxxxxxxxxssssssssxxxxxxxxxxxxxx
17 Fast Internally Circulating Fluidised Bed
Figure 21:Scheme of MILENA gasifier
-
43
The combustor operates as a Bubbling Fluidized Bed. The downcomers
transport bed material and char from the gasification section into the
combustion. Tar and separated from the producer can also be fed into the
combustor. Char, tar and dust are burned with air to heat the bed material to
around 925⁰C. No additional heat input is required since all heat for the
gasification process is produced by the combustion of the char, tar and dust.
A typical MILENA gas composition, on dry basis, after tar removal, has been
shown previously in Figure 2. Because steam was used as gasifier agent, the
product gas contains only a small amount of N2. As stated before MILENA
gasifier was designed as a first step for the production of SNG. Therefore the
high yield of CH4 is very attractive.
MILENA gasifier had not been modelled in AspenPlus. Compositions and
conditions of inlet and outlet streams from gasification sector were obtained
from pilot plant results.
A.2.2 OLGA tar removal (all cases)
Tars can be considered as major technical obstacle in the implementation of
biomass gasification technology. Condensing tars dramatically foul the
downstream system piping and gas cooling and cleaning equipment, while
liquid tar droplets (i.e. aerosols) that enter prime movers disturb the operation
of these end-use applications of the product gas. Tar plays also negative role in
the wastewater treatment problem. Removal of tar components wastewater
requires considerable investments that can even be dramatic as some tar
components show poisoning behaviour in biologic wastewater treatment
systems (e.g. phenol).
Indirect combustion produces a combustible gas containing tar. Main reason for
tars to survive is the low fluidized bed gasifier temperature (
-
44
applied for this purpose. Additional benefit is that hydrogen chloride will be
removed as well.
NH3 removal has been modelled in AspenPlus as separator block. 99.99% level
of separation is assumed for both NH3 and HCl.
A.2.4 EB reactor (EB case)
The reactor system is comprised of two parallel fixed bed reactors. Zeolite ZSM-
5 is used as catalyst, which has a low coking tendency and, therefore, life cycles
between regenerations up to 120 days are expected. This was shown on the
example on non-purified FCC off-gas, which was used as a diluted ethylene
feedstock [5]. Catalyst activity can be renewed with when the deposed coke is
burned off in air stream.
The kinetic data for alkylation of benzene on ZSM-5 found in literature [19]
were tested with the input of relevant research papers [15]. Four sets of kinetic
data were tested but results were unsatisfactory. It was concluded that kinetic
data from literature are not suitable for extrapolation at different conditions.
Therefore, alkylation reactor has been modelled as R-STOIC reactor. Conversion
of ethylene and selectivity towards ethylbenzene are assumed to be 95% and
90% respectively. This assumption was based on research paper from Graf [15].
The ratio of ethylene and benzene in the reactor feed was set to be 5.
Pressure drop is assumed to be 1 bar.
A.2.5 Separation section (EB case)
Separation section in the EB case consists of flash drum, distillation column and
BTX scrubber (see appendix A.4.1). Before entering the distillation column the
major part of the gases is removed by flash drum. Vapour stream from the flash
drum, since it still contains a large part of the benzene, is sent to BTX scrubber
for the recovery of benzene.
Liquid stream from the flash drum is fed to the distillation column. Since
relative volatility between benzene and ethylbenzene decreases with higher
pressures and temperatures, pressure in distillation column is set to 1 bar.
Operating the distillation column at atmospheric pressure also decreases the
temperature in the reboilier section, which implies that lower pressure steam
can be used as utility. In the condenser section, cooling water of 20°C is used as
coolant. Distillation column has 13 trays and feed at stage 9 is found as an
optimum.
Top fraction of the distillation column (99% benzene) is mixed with the benzene
recovered from the BTX scrubber and subsequently sent to evaporator. Vapour
benzene is then sent back to the EB reactor.
A.2.6 DEA reactor (DEA-1, DEA-2 case)
-
45
Diluted ethylene aromatization reactor is also considered to be system of two
parallel fixed bed reactors. Catalyst used would be ZSM5 DEA-1, or ZSM
impregnated with Gallium in DEA-2.
It is unclear how much would be life cycles between the regeneration.
Durability tests are needed to be performed for that purpose. It is assumed that
catalyst activity can be fully restored after burning off the coke in the air
stream.
No kinetic data was found in literature for aromatization of ethylene.
Therefore, DEA reactor was modelled in AspenPlus as R-STOIC reactor with
following assumptions:
DEA-1 case:
Conversion of ethylene is assumed to be 81%, based on experimental results (see Chapter 3.3)
Selectivity to aromatics is assumed to be 40% based on experimental results (see Chapter 3.3)
Acetylene is assumed to convert partially to ethylene and partially to naphthalene
DEA-2 case:
Conversion of ethylene is assumed to be 95%, based on literature [8] Selectivity to aromatics is assumed to be 80%
Distribution between aromatics is assumed to be the same as in lab tests (See Chapter 3.3 Figure 9)
Acetylene is assumed to convert partially to ethylene and partially to
naphthalene
A.2.7 BTX removal/Aromatics removal (BTX, EB, DEA-
1, DEA-2 case)
ECN recently started a development of technology for separation of bioBTX
from the producer gas.
The absorption was chosen as best technology for that purpose. Currently,
there are lab scale absorber and striper columns with 2m height. Dioxin was
chosen as a first choice for absorber liquid.
However, dioxin that were used had problem with solidification on temperature
around 50°C and issues with thermal stability [2].Recent development of BTX
scrubber managed to overcome these difficulties, by using silica based
scrubbing oil. These improvements made system also suitable for separation of
higher aromatics (DEA case), due to the higher temperature needed in the
striping section. However, additional research and development of BTX
scrubber still needs to be done.
BTX scrubber has been modelled in AspenPlus as separation block. It was
assumed 99.9% removal with oil striper, and rest removed with activated
-
46
carbon guard bed. Thiophene is also considered to be removed in the BTX